Job Description
The company and our mission: Zartis is a global AI transformation and technology consulting partner where talented engineers and technologists work on cutting-edge innovation. We partner with ambitious organizations to design, build, and scale technology solutions that deliver real impact. Our teams bring deep expertise in AI-driven platforms, secure API architectures, and cloud-native engineering. You will work on meaningful projects that accelerate the adoption of advanced technologies, from strategy and discovery through to full product delivery, helping turn complex challenges into measurable outcomes. With engineering hubs across EMEA and LATAM, and long-term partnerships in financial services, healthcare and life sciences, and energy and climate, we offer opportunities to work on projects that truly matter. Here, you will not just build technology — you will drive business impact and grow your career alongside industry leaders. We are looking for a Principal DevOps Engineer to work on a project in the marketplace sector. The project: We are looking for a talented Principal DevOps Engineer to join a founding-stage initiative at one of Europe's largest digital-native groups, building the internal agentic platform that enables safe, observable, and reusable AI-driven workflows at marketplace scale. You will be part of a founding platform cell at the heart of an AI House, the central engine for making agentic engineering the default across all marketplace teams. The platform does not exist yet: you help define what it is. You will work across secure agent runtimes, tool gateways, evaluation pipelines, observability systems, and reusable scaffolding, all built for a heterogeneous, multi-cloud, multi-stack estate spanning products like Leboncoin, Kleinanzeigen, Marktplaats, Subito, Mobile.de, and more. Our teammates come from a variety of backgrounds and we are committed to building an inclusive culture based on trust and innovation. What you will do: Build and operate secure agent runtimes with sandboxing, runtime isolation, network policies, secrets management, RBAC, and approval flows that make agents a credible part of the infrastructure. Design and maintain the integration surface with MCP-style adapters and gateways that let agents act on source control, CI/CD, ticketing, documentation, cloud, and observability systems across all marketplace teams. Build and own the evaluation pipeline with golden tasks, graders, regression tests, and release gates that make agentic workflow correctness measurable and cheap enough that teams actually use it. Implement observability and cost control including traces, telemetry, token usage, cost-per-workflow, rate limits, and failure handling so stakeholders always know what is running, what it costs, and what to do when it breaks. Create reusable scaffolding including templates, starter kits, wrapper scripts, and workflow automation that turn proven patterns from marketplace pilots into platform components every team can adopt in a day. Partner with the AI Architect to make the reference architecture real, in code, in production, on a schedule that does not slip. What you will bring: A track record of shipping internal developer platforms or developer tooling that real teams depend on, with users, SLAs, and on-call rotations, not just side projects. Hands-on experience designing and operating secure execution environments including sandboxing, runtime isolation, RBAC, secrets, and audit logging at production scale. A history of shipping integrations across source control, CI/CD, ticketing, documentation, cloud, and observability systems across polyglot estates. Experience building or operating tool gateways, adapters, or MCP-style integrations for agents and an understanding of the new failure modes they introduce. An instrumentation-first mindset where traces, telemetry, cost-per-workflow, and eval pipelines are standard practice, not afterthoughts. An SRE mindset where reliability, rollback, rate limiting, graceful degradation, and cost control are first-class concerns from day one. A record of shipping reusable templates and starter kits that teams actually adopt rather than internal frameworks that gather dust. Comfort across modern infrastructure including containers, cloud (AWS or equivalent), IaC, CI/CD, APIs, and scripting languages. Experience supporting many engineers or multiple teams as internal customers, balancing opinionated defaults with inevitable special cases. The ability to collaborate closely with security, legal, and risk stakeholders without losing momentum. Nice to have: Prior experience building agentic platforms or AI infrastructure at multi-team scale. Familiarity with MCP (Model Context Protocol) integrations or equivalent agent tool gateway patterns. Experience with multi-cloud or multi-stack platform engineering across distributed marketplace or e-commerce environments. Exposure to cost governance tooling, token budgeting, or LLM observab
